Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/niconico.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import datetime
   6
   7 from .common import InfoExtractor
   8 from ..compat import (
   9     compat_parse_qs,
  10     compat_urlparse,
  11 )
  12 from ..utils import (
  13     determine_ext,
  14     dict_get,
  15     ExtractorError,
  16     int_or_none,
  17     float_or_none,
  18     parse_duration,
  19     parse_iso8601,
  20     remove_start,
  21     try_get,
  22     unified_timestamp,
  23     urlencode_postdata,
  24     xpath_text,
  25 )
  26
  27
  28 class NiconicoIE(InfoExtractor):
  29     IE_NAME = 'niconico'
  30     IE_DESC = 'ニコニコ動画'
  31
  32     _TESTS = [{
  33         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  34         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  35         'info_dict': {
  36             'id': 'sm22312215',
  37             'ext': 'mp4',
  38             'title': 'Big Buck Bunny',
  39             'thumbnail': r're:https?://.*',
  40             'uploader': 'takuya0301',
  41             'uploader_id': '2698420',
  42             'upload_date': '20131123',
  43             'timestamp': int,  # timestamp is unstable
  44             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  45             'duration': 33,
  46             'view_count': int,
  47             'comment_count': int,
  48         },
  49         'skip': 'Requires an account',
  50     }, {
  51         # File downloaded with and without credentials are different, so omit
  52         # the md5 field
  53         'url': 'http://www.nicovideo.jp/watch/nm14296458',
  54         'info_dict': {
  55             'id': 'nm14296458',
  56             'ext': 'swf',
  57             'title': '【鏡音リン】Dance on media【オリジナル】take2!',
  58             'description': 'md5:689f066d74610b3b22e0f1739add0f58',
  59             'thumbnail': r're:https?://.*',
  60             'uploader': 'りょうた',
  61             'uploader_id': '18822557',
  62             'upload_date': '20110429',
  63             'timestamp': 1304065916,
  64             'duration': 209,
  65         },
  66         'skip': 'Requires an account',
  67     }, {
  68         # 'video exists but is marked as "deleted"
  69         # md5 is unstable
  70         'url': 'http://www.nicovideo.jp/watch/sm10000',
  71         'info_dict': {
  72             'id': 'sm10000',
  73             'ext': 'unknown_video',
  74             'description': 'deleted',
  75             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  76             'thumbnail': r're:https?://.*',
  77             'upload_date': '20071224',
  78             'timestamp': int,  # timestamp field has different value if logged in
  79             'duration': 304,
  80             'view_count': int,
  81         },
  82         'skip': 'Requires an account',
  83     }, {
  84         'url': 'http://www.nicovideo.jp/watch/so22543406',
  85         'info_dict': {
  86             'id': '1388129933',
  87             'ext': 'mp4',
  88             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
  89             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
  90             'thumbnail': r're:https?://.*',
  91             'timestamp': 1388851200,
  92             'upload_date': '20140104',
  93             'uploader': 'アニメロチャンネル',
  94             'uploader_id': '312',
  95         },
  96         'skip': 'The viewing period of the video you were searching for has expired.',
  97     }, {
  98         # video not available via `getflv`; "old" HTML5 video
  99         'url': 'http://www.nicovideo.jp/watch/sm1151009',
 100         'md5': '8fa81c364eb619d4085354eab075598a',
 101         'info_dict': {
 102             'id': 'sm1151009',
 103             'ext': 'mp4',
 104             'title': 'マスターシステム本体内蔵のスペハリのメインテーマ（ＰＳＧ版）',
 105             'description': 'md5:6ee077e0581ff5019773e2e714cdd0b7',
 106             'thumbnail': r're:https?://.*',
 107             'duration': 184,
 108             'timestamp': 1190868283,
 109             'upload_date': '20070927',
 110             'uploader': 'denden2',
 111             'uploader_id': '1392194',
 112             'view_count': int,
 113             'comment_count': int,
 114         },
 115         'skip': 'Requires an account',
 116     }, {
 117         # "New" HTML5 video
 118         # md5 is unstable
 119         'url': 'http://www.nicovideo.jp/watch/sm31464864',
 120         'info_dict': {
 121             'id': 'sm31464864',
 122             'ext': 'mp4',
 123             'title': '新作TVアニメ「戦姫絶唱シンフォギアAXZ」PV 最高画質',
 124             'description': 'md5:e52974af9a96e739196b2c1ca72b5feb',
 125             'timestamp': 1498514060,
 126             'upload_date': '20170626',
 127             'uploader': 'ゲスト',
 128             'uploader_id': '40826363',
 129             'thumbnail': r're:https?://.*',
 130             'duration': 198,
 131             'view_count': int,
 132             'comment_count': int,
 133         },
 134         'skip': 'Requires an account',
 135     }, {
 136         # Video without owner
 137         'url': 'http://www.nicovideo.jp/watch/sm18238488',
 138         'md5': 'd265680a1f92bdcbbd2a507fc9e78a9e',
 139         'info_dict': {
 140             'id': 'sm18238488',
 141             'ext': 'mp4',
 142             'title': '【実写版】ミュータントタートルズ',
 143             'description': 'md5:15df8988e47a86f9e978af2064bf6d8e',
 144             'timestamp': 1341160408,
 145             'upload_date': '20120701',
 146             'uploader': None,
 147             'uploader_id': None,
 148             'thumbnail': r're:https?://.*',
 149             'duration': 5271,
 150             'view_count': int,
 151             'comment_count': int,
 152         },
 153         'skip': 'Requires an account',
 154     }, {
 155         'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg',
 156         'only_matching': True,
 157     }]
 158
 159     _VALID_URL = r'https?://(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch/(?P<id>(?:[a-z]{2})?[0-9]+)'
 160     _NETRC_MACHINE = 'niconico'
 161
 162     def _real_initialize(self):
 163         self._login()
 164
 165     def _login(self):
 166         username, password = self._get_login_info()
 167         # No authentication to be performed
 168         if not username:
 169             return True
 170
 171         # Log in
 172         login_ok = True
 173         login_form_strs = {
 174             'mail_tel': username,
 175             'password': password,
 176         }
 177         urlh = self._request_webpage(
 178             'https://account.nicovideo.jp/api/v1/login', None,
 179             note='Logging in', errnote='Unable to log in',
 180             data=urlencode_postdata(login_form_strs))
 181         if urlh is False:
 182             login_ok = False
 183         else:
 184             parts = compat_urlparse.urlparse(urlh.geturl())
 185             if compat_parse_qs(parts.query).get('message', [None])[0] == 'cant_login':
 186                 login_ok = False
 187         if not login_ok:
 188             self._downloader.report_warning('unable to log in: bad username or password')
 189         return login_ok
 190
 191     def _extract_format_for_quality(self, api_data, video_id, audio_quality, video_quality):
 192         def yesno(boolean):
 193             return 'yes' if boolean else 'no'
 194
 195         session_api_data = api_data['video']['dmcInfo']['session_api']
 196         session_api_endpoint = session_api_data['urls'][0]
 197
 198         format_id = '-'.join(map(lambda s: remove_start(s['id'], 'archive_'), [video_quality, audio_quality]))
 199
 200         session_response = self._download_json(
 201             session_api_endpoint['url'], video_id,
 202             query={'_format': 'json'},
 203             headers={'Content-Type': 'application/json'},
 204             note='Downloading JSON metadata for %s' % format_id,
 205             data=json.dumps({
 206                 'session': {
 207                     'client_info': {
 208                         'player_id': session_api_data['player_id'],
 209                     },
 210                     'content_auth': {
 211                         'auth_type': session_api_data['auth_types'][session_api_data['protocols'][0]],
 212                         'content_key_timeout': session_api_data['content_key_timeout'],
 213                         'service_id': 'nicovideo',
 214                         'service_user_id': session_api_data['service_user_id']
 215                     },
 216                     'content_id': session_api_data['content_id'],
 217                     'content_src_id_sets': [{
 218                         'content_src_ids': [{
 219                             'src_id_to_mux': {
 220                                 'audio_src_ids': [audio_quality['id']],
 221                                 'video_src_ids': [video_quality['id']],
 222                             }
 223                         }]
 224                     }],
 225                     'content_type': 'movie',
 226                     'content_uri': '',
 227                     'keep_method': {
 228                         'heartbeat': {
 229                             'lifetime': session_api_data['heartbeat_lifetime']
 230                         }
 231                     },
 232                     'priority': session_api_data['priority'],
 233                     'protocol': {
 234                         'name': 'http',
 235                         'parameters': {
 236                             'http_parameters': {
 237                                 'parameters': {
 238                                     'http_output_download_parameters': {
 239                                         'use_ssl': yesno(session_api_endpoint['is_ssl']),
 240                                         'use_well_known_port': yesno(session_api_endpoint['is_well_known_port']),
 241                                     }
 242                                 }
 243                             }
 244                         }
 245                     },
 246                     'recipe_id': session_api_data['recipe_id'],
 247                     'session_operation_auth': {
 248                         'session_operation_auth_by_signature': {
 249                             'signature': session_api_data['signature'],
 250                             'token': session_api_data['token'],
 251                         }
 252                     },
 253                     'timing_constraint': 'unlimited'
 254                 }
 255             }).encode())
 256
 257         resolution = video_quality.get('resolution', {})
 258
 259         return {
 260             'url': session_response['data']['session']['content_uri'],
 261             'format_id': format_id,
 262             'ext': 'mp4',  # Session API are used in HTML5, which always serves mp4
 263             'abr': float_or_none(audio_quality.get('bitrate'), 1000),
 264             'vbr': float_or_none(video_quality.get('bitrate'), 1000),
 265             'height': resolution.get('height'),
 266             'width': resolution.get('width'),
 267         }
 268
 269     def _real_extract(self, url):
 270         video_id = self._match_id(url)
 271
 272         # Get video webpage. We are not actually interested in it for normal
 273         # cases, but need the cookies in order to be able to download the
 274         # info webpage
 275         webpage, handle = self._download_webpage_handle(
 276             'http://www.nicovideo.jp/watch/' + video_id, video_id)
 277         if video_id.startswith('so'):
 278             video_id = self._match_id(handle.geturl())
 279
 280         api_data = self._parse_json(self._html_search_regex(
 281             'data-api-data="([^"]+)"', webpage,
 282             'API data', default='{}'), video_id)
 283
 284         def _format_id_from_url(video_url):
 285             return 'economy' if video_real_url.endswith('low') else 'normal'
 286
 287         try:
 288             video_real_url = api_data['video']['smileInfo']['url']
 289         except KeyError:  # Flash videos
 290             # Get flv info
 291             flv_info_webpage = self._download_webpage(
 292                 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
 293                 video_id, 'Downloading flv info')
 294
 295             flv_info = compat_urlparse.parse_qs(flv_info_webpage)
 296             if 'url' not in flv_info:
 297                 if 'deleted' in flv_info:
 298                     raise ExtractorError('The video has been deleted.',
 299                                          expected=True)
 300                 elif 'closed' in flv_info:
 301                     raise ExtractorError('Niconico videos now require logging in',
 302                                          expected=True)
 303                 elif 'error' in flv_info:
 304                     raise ExtractorError('%s reports error: %s' % (
 305                         self.IE_NAME, flv_info['error'][0]), expected=True)
 306                 else:
 307                     raise ExtractorError('Unable to find video URL')
 308
 309             video_info_xml = self._download_xml(
 310                 'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id,
 311                 video_id, note='Downloading video info page')
 312
 313             def get_video_info(items):
 314                 if not isinstance(items, list):
 315                     items = [items]
 316                 for item in items:
 317                     ret = xpath_text(video_info_xml, './/' + item)
 318                     if ret:
 319                         return ret
 320
 321             video_real_url = flv_info['url'][0]
 322
 323             extension = get_video_info('movie_type')
 324             if not extension:
 325                 extension = determine_ext(video_real_url)
 326
 327             formats = [{
 328                 'url': video_real_url,
 329                 'ext': extension,
 330                 'format_id': _format_id_from_url(video_real_url),
 331             }]
 332         else:
 333             formats = []
 334
 335             dmc_info = api_data['video'].get('dmcInfo')
 336             if dmc_info:  # "New" HTML5 videos
 337                 quality_info = dmc_info['quality']
 338                 for audio_quality in quality_info['audios']:
 339                     for video_quality in quality_info['videos']:
 340                         if not audio_quality['available'] or not video_quality['available']:
 341                             continue
 342                         formats.append(self._extract_format_for_quality(
 343                             api_data, video_id, audio_quality, video_quality))
 344
 345                 self._sort_formats(formats)
 346             else:  # "Old" HTML5 videos
 347                 formats = [{
 348                     'url': video_real_url,
 349                     'ext': 'mp4',
 350                     'format_id': _format_id_from_url(video_real_url),
 351                 }]
 352
 353             def get_video_info(items):
 354                 return dict_get(api_data['video'], items)
 355
 356         # Start extracting information
 357         title = get_video_info('title')
 358         if not title:
 359             title = self._og_search_title(webpage, default=None)
 360         if not title:
 361             title = self._html_search_regex(
 362                 r'<span[^>]+class="videoHeaderTitle"[^>]*>([^<]+)</span>',
 363                 webpage, 'video title')
 364
 365         watch_api_data_string = self._html_search_regex(
 366             r'<div[^>]+id="watchAPIDataContainer"[^>]+>([^<]+)</div>',
 367             webpage, 'watch api data', default=None)
 368         watch_api_data = self._parse_json(watch_api_data_string, video_id) if watch_api_data_string else {}
 369         video_detail = watch_api_data.get('videoDetail', {})
 370
 371         thumbnail = (
 372             get_video_info(['thumbnail_url', 'thumbnailURL']) or
 373             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
 374             video_detail.get('thumbnail'))
 375
 376         description = get_video_info('description')
 377
 378         timestamp = (parse_iso8601(get_video_info('first_retrieve')) or
 379                      unified_timestamp(get_video_info('postedDateTime')))
 380         if not timestamp:
 381             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
 382             if match:
 383                 timestamp = parse_iso8601(match.replace('+', ':00+'))
 384         if not timestamp and video_detail.get('postedAt'):
 385             timestamp = parse_iso8601(
 386                 video_detail['postedAt'].replace('/', '-'),
 387                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 388
 389         view_count = int_or_none(get_video_info(['view_counter', 'viewCount']))
 390         if not view_count:
 391             match = self._html_search_regex(
 392                 r'>Views: <strong[^>]*>([^<]+)</strong>',
 393                 webpage, 'view count', default=None)
 394             if match:
 395                 view_count = int_or_none(match.replace(',', ''))
 396         view_count = view_count or video_detail.get('viewCount')
 397
 398         comment_count = (int_or_none(get_video_info('comment_num')) or
 399                          video_detail.get('commentCount') or
 400                          try_get(api_data, lambda x: x['thread']['commentCount']))
 401         if not comment_count:
 402             match = self._html_search_regex(
 403                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
 404                 webpage, 'comment count', default=None)
 405             if match:
 406                 comment_count = int_or_none(match.replace(',', ''))
 407
 408         duration = (parse_duration(
 409             get_video_info('length') or
 410             self._html_search_meta(
 411                 'video:duration', webpage, 'video duration', default=None)) or
 412             video_detail.get('length') or
 413             get_video_info('duration'))
 414
 415         webpage_url = get_video_info('watch_url') or url
 416
 417         # Note: cannot use api_data.get('owner', {}) because owner may be set to "null"
 418         # in the JSON, which will cause None to be returned instead of {}.
 419         owner = try_get(api_data, lambda x: x.get('owner'), dict) or {}
 420         uploader_id = get_video_info(['ch_id', 'user_id']) or owner.get('id')
 421         uploader = get_video_info(['ch_name', 'user_nickname']) or owner.get('nickname')
 422
 423         return {
 424             'id': video_id,
 425             'title': title,
 426             'formats': formats,
 427             'thumbnail': thumbnail,
 428             'description': description,
 429             'uploader': uploader,
 430             'timestamp': timestamp,
 431             'uploader_id': uploader_id,
 432             'view_count': view_count,
 433             'comment_count': comment_count,
 434             'duration': duration,
 435             'webpage_url': webpage_url,
 436         }
 437
 438
 439 class NiconicoPlaylistIE(InfoExtractor):
 440     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/mylist/(?P<id>\d+)'
 441
 442     _TEST = {
 443         'url': 'http://www.nicovideo.jp/mylist/27411728',
 444         'info_dict': {
 445             'id': '27411728',
 446             'title': 'AKB48のオールナイトニッポン',
 447         },
 448         'playlist_mincount': 225,
 449     }
 450
 451     def _real_extract(self, url):
 452         list_id = self._match_id(url)
 453         webpage = self._download_webpage(url, list_id)
 454
 455         entries_json = self._search_regex(r'Mylist\.preload\(\d+, (\[.*\])\);',
 456                                           webpage, 'entries')
 457         entries = json.loads(entries_json)
 458         entries = [{
 459             '_type': 'url',
 460             'ie_key': NiconicoIE.ie_key(),
 461             'url': ('http://www.nicovideo.jp/watch/%s' %
 462                     entry['item_data']['video_id']),
 463         } for entry in entries]
 464
 465         return {
 466             '_type': 'playlist',
 467             'title': self._search_regex(r'\s+name: "(.*?)"', webpage, 'title'),
 468             'id': list_id,
 469             'entries': entries,
 470         }